Finding high-redshift strong lenses in DES using convolutional neural networks

Author(s): Jacobs, C; Collett, T; Glazebrook, K; McCarthy, C; Qin, AK; Abbott, TMC; Abdalla, FB; Annis, J; Avila, S; Bechtol, K; Bertin, E; Brooks, D; Buckley-Geer, E; Burke, DL; Carnero Rosell, A; Carrasco Kind, M; Carretero, J; Da Costa, LN; Davis, C; De Vicente, J; Desai, S; Diehl, HT; Doel, P; Eifler, TF; Flaugher, B; Frieman, J; Garcia-Bellido, J; Gaztanaga, E; Gerdes, DW; Goldstein, DA; Gruen, D; Gruendl, RA; Gschwend, J; Gutierrez, G; Hartley, WG; Hollowood, DL; Honscheid, K; Hoyle, B; James, DJ; Kuehn, K; Kuropatkin, N; Lahav, O; Li, TS; Lima, M; Lin, H; Maia, MAG; Martini, P; Miller, CJ; Miquel, R; Nord, B; Plazas, AA; Sanchez, E; Scarpine, V; Schubnell, M; Serrano, S; Sevilla-Noarbe, I; Smith, M; Soares-Santos, M; Sobreira, F; Suchyta, E; Swanson, MEC; Tarle, G; Vikram, V; Walker, AR; Zhang, Y; Zuntz, J | Abstract: © 2019 The Author(s) Published by Oxford University Press on behalf of the Royal Astronomical Society. We search Dark Energy Survey (DES) Year 3 imaging data for galaxy-galaxy strong gravitational lenses using convolutional neural networks. We generate 250 000 simulated lenses at redshifts g 0.8 from which we create a data set for training the neural networks with realistic seeing, sky and shot noise. Using the simulations as a guide, we build a catalogue of 1.1 million DES sources with 1.8 l g − i l 5, 0.6 l g − r l 3, r mag g 19, g mag g 20, and i mag g 18.2. We train two ensembles of neural networks on training sets consisting of simulated lenses, simulated non-lenses, and real sources. We use the neural networks to score images of each of the sources in our catalogue with a value from 0 to 1, and select those with scores greater than a chosen threshold for visual inspection, resulting in a candidate set of 7301 galaxies. During visual inspection, we rate 84 as 'probably' or 'definitely' lenses. Four of these are previously known lenses or lens candidates. We inspect a further 9428 candidates with a different score threshold, and identify four new candidates. We present 84 new strong lens candidates, selected after a few hours of visual inspection by astronomers. This catalogue contains a comparable number of high-redshift lenses to that predicted by simulations. Based on simulations, we estimate our sample to contain most discoverable lenses in this imaging and at this redshift range.